[almostgotcaught]

> this repo massively oversells itself

that's literally every single paper in academic CS (except theory). every claim is an exaggeration or lie by omission ("this only works on a small subset of ..."). every title is clickbait. every impl is worse than junior-level code. you become so jaded to it that you start to evaluate papers/projects based a premonition of what it could be rather than what it is.